Photovoltaic Energy Conversion System Fault Detection Using Fractional-Order Color Relation Classifier in Microdistribution Systems

In this paper, we propose a photovoltaic (PV) energy conversion system (PVECS) fault detection scheme using a fractional-order color relation classifier in microdistribution systems. Based on electrical examination method, output power degradation is used to monitor physical conditions with changes in a PV array's circuitry, including grounded faults, mismatch faults, bridged faults between two PV panels, and open-circuit faults. The PV array power depends on solar radiation and temperature, and maximum power point tracking (MPPT) control is used to maintain stable power supply to a microdistribution system in the event of a fault in the PVECS. The MPPT algorithm is employed to estimate the desired maximum power, which is then compared with the meter-read power. Fractional-order dynamic errors are determined to quantify output power degradation between the desired maximum power and the meter-read power. Then, a color relation analysis is used to separate normal conditions from fault events. For a PVECS with two panels in parallel, the simulation results demonstrate that the proposed method is suitable for real-time applications and is flexible for fault identification. Its detection rates exceeded 88.23% for six events.